Facilitate publishing a major version

[peterdesmet]

When republishing a dataset, the following text is presented:

You are about to publish a new minor version of this resource. If there have been scientifically significant changes to the resource since the last publication, you should cancel, assign this resource a new DOI, and then publish again.

I think this is somewhat unhelpful:

1. What is a scientifically significant change?
2. How do I assign a new DOI (if my resources are assigned a DOI via GBIF)?

Note: there is probably a larger discussion needed on DOIs and versioning.

• Should the data publisher decide if a version is minor or major or should that be automatically assigned by the IPT based on certain criteria (e.g. metadata update vs data update)?
• Should it be possible for a data publisher to edit/update metadata without creating a new (minor) version (cf. Zenodo)?
• Should there be a version-less DOI that always points to the most current one (cf. Zenodo)?

In the meantime, I think it would be good to update the help text to:

You are about to publish a new minor version of this resource. If there have been scientifically significant changes to the resource since the last publication and this IPT is configured to assign DOIs, you should cancel, assign this resource a new DOI, and then publish again.

[MattBlissett]

There is a larger discussion needed around this -- some people treat the IPT as a full data repository, for others it is just the means to get data to GBIF.

Either way, the IPT knows if the current resource has an IPT-managed DOI, so I think we can keep the current message in that case, and add a new, simpler message otherwise:

You are about to publish a new minor version of this resource. If there have been scientifically significant changes to the resource since the last publication you should cancel, and publish a new major version.

We'd then need a way to publish a new, major version. (Or maybe the versions are irrelevant unless the IPT has assigned DOIs.)

[peterdesmet]

Yeah, I find the current message confusing for an IPT that is not set up with DOIs, so I agree with:

Or maybe the versions are irrelevant unless the IPT has assigned DOIs

For IPTs that do not assign DOIs, I would write:

You are about to publish a new minor version of this resource.

[abubelinha]

For IPTs that do not assign DOIs, I would write:

You are about to publish a new minor version of this resource.

Yes, I agree.
All current DOIs wording is pretty unhelpful if the IPT does not assign DOIs.

[peterdesmet]

@mike-podolskiy90 note that the same phrase is used in the dialog box when republishing a Data Package. The DOI mention is confusing here too, and "You are about to publish a new public minor version of this resource." is incorrect, since Data Packages are published as major releases (1.0 -> 2.0 -> 3.0).

[mike-podolskiy90]

We simplified the message for Data package resources (there is no minor/major versions there, #2293 ), but it's still same confusing for DwC.

So we need to decide what to do.

[CecSve]

Relevant Discourse thread on the topic: https://discourse.gbif.org/t/when-to-assign-a-new-doi-to-an-existing-dataset/2319/3

[CecSve]

Maybe this would be relevant here as well?:
#1900 (comment)